Loading
Ace Intelligence Systems
Preparing a calmer, clearer view of your automation workspace.
Aligning workflows
Loading
Preparing a calmer, clearer view of your automation workspace.
Ace Intelligence
An advanced autonomous multi-agent research platform powered by n8n orchestration. 8 specialized AI agents collaborate to transform a research topic into a comprehensive academic paper.
Version 8.0 — Production-Ready, Single Unified Workflow. This system represents a sophisticated deep-engineering build that automates the lifecycle of complex research. It relies on LangGraph principles to orchestrate multi-agent workflows, featuring specialized discovery agents, systematic literature review capabilities, and strict hallucination-reducing verification layers. The modular pipeline integrates external APIs like arXiv and Semantic Scholar alongside ChromaDB for robust vector memory and RAG.
The system runs as a unified n8n workflow. Submit a research topic via the frontend, and 8 agents autonomously execute the full research lifecycle.
Orchestrator: State initialization and management via Code Node. Keyword Generator: Academic search keyword generation using LLaMA 3.3-70B. Researcher: Literature search and discovery via HTTP/API. Strategist: Gap identification and research strategy using LLaMA 3.3-70B. Architect: Methodology design using LLaMA 3.3-70B. Implementer: Data and implementation planning using LLaMA 3.3-70B. Analyst: Experiment design using LLaMA 3.3-70B. Editor: Final paper compilation using LLaMA 3.3-70B.
Phase 1 — Initialization: Webhook trigger receives topic from frontend, initializes research state. Phase 2 — Research Intelligence (Agents 1-3): Keyword generation, web search via DuckDuckGo, literature review with theme identification. Phase 3 — Strategy & Methodology (Agents 4-5): Gap statement formulation, research questions, methodology design. Phase 4 — Implementation & Experiments (Agents 6-7): Data requirements planning, experimental framework design. Phase 5 — Quality & Compilation (Agents 7-8): Novelty and ethics validation, IEEE-format paper generation. Phase 6 — Response: JSON-formatted output with complete paper and execution metadata.
Modern glassmorphism design with real-time agent progress visualization. Markdown rendering with syntax highlighting, copy and download functionality, and fully responsive design.
Endpoint: POST http://localhost:5678/webhook/start-research. Request body: { 'topic': 'Your Research Topic' }. Response includes the full generated paper content, execution time, agents executed, and phase completion status.
POST /webhook/start-research
Request: { "topic": "Your Topic" }
Response: { "success": true, "content": "# Paper...", "metadata": { "executionTimeSeconds": 90, "agentsExecuted": 8 } }This project serves as our ultimate proof of technical capability. While rapid-deployment automations act as the 'Trojan Horse' for initial trust, this system is the high-ticket upsell. It proves to enterprise organizations that we can move beyond simple chatbots to deploy autonomous agents that research, reason, verify facts, and execute multi-step tasks securely at scale. Falls under our Custom Generative AI & Conversational Agents category — a flagship example for Multi-Agent Workflows and Enterprise RAG Architectures.
Explore the full n8n workflow JSON, frontend code, and deployment configuration.
https://github.com/OMCHOKSI108/AI-AUTOMATION-WORKFLOWS